摘要:
Sciarapotamo Bridge located in Reggio Calabria Province, Italy has been recently retrofitted by replacing the deteriorated concrete bridge deck with composite concrete-steel deck, and isolating the deck with eight high damping rubber (HDR) and four multi-directional sliding bearings. Full-scale ambient vibration dynamic commissioning tests were performed on the bridge in June 2012 just before the bridge has become fully operational. The dynamic response of the bridge was recorded using an array of 4 uni-axial and 4 tri-axial force-balanced accelerometers placed along the entire length of the bridge. Different output-only system identification methods are used to identify the modal parameters of the bridge including: (1) Multiple Reference Natural Excitation Technique in conjunction with Eigensystem Realization Algorithm (MNExT-ERA), (2) Enhanced Frequency Domain Decomposition (EFDD). In this study, modal parameter estimation results obtained using different system identification methods are compared. Especially estimated damping coefficients are of great value showing the total damping in the system in as-built conditions. The experimentally obtained modal parameters for the bridge set the undamaged benchmark state of the bridge for future potential damage assessment studies. 1 INTRODUCTION Civil engineering structures are exposed to different external effects that change their dynamic characteristics. Damage can be defined as changes affecting the structural performance of a system. The fact that damage can alter stiffness, mass, and/or energy dissipation capacity of a structure, which in turn results in detectable changes in its vibration signature, is the underlying principle of vibration-based structural health monitoring (SHM) (Doebling et al., 1998). SHM involves monitoring a structure continuously or intermittently and extracting damage sensitive properties for assessing its current state. Estimating dynamic parameters (i.e., modal parameters) by using global vibration response has attracted increasing attention in recent years in civil engineering research community, and has become an important tool for civil engineering structures for damage assessment and prognosis, model calibration, assessing retrofitting strategy and performance before and after retrofitting work (Farrar et al., 2007; Nayeri et al., 2007). Experimental modal analysis is used as a technology to extract modal parameters of vibrating structures using low level vibration data. Modal parameters to be extracted are natural frequencies, mode shapes, damping ratios, and modal participation factors (Peeters et al., 2001). It is possible to classify system identification methods used in structural health monitoring in two categories, input-output methods and only-output methods. Exciting civil engineering
展开